PUBLIC TRUST AND CONFIDENCE
Proposal:
72. The consultation paper "Improving Statistical Services in Scotland" proposed to encourage Parliamentary interest in statistics and to make it clear that a Committee is welcome to question statisticians directly. It suggested a range of measure to increase confidence and to stimulate debate:
- separately brand the material that has been produced by the professionally independent statistics staff from Ministerial statements;
- make further improvements to access to statistics, background data, methods used and supporting information to allow users to make their own assessment of the reliability of our statistics;
- engage with key commentators on statistics, including making a baseline assessment of their confidence in official statistics;
- adopt a more pro-active approach with individuals who comment or criticise official statistics;
- develop information on official statistics to be used in schools; and
- develop the existing ScotStat network, including considering general analytical services issues as well as technical statistics matters.
And the consultation paper asked:
- Are these proposals useful in increasing public confidence in statistics?
- Should more be done?
Findings:
Consultation Paper
73. Consultation paper responses showed support for the proposals, and other suggestions were made such as peer reviews and more promotion of ScotStat.
- 'Scottish Water would also like to see the ScotStat group and website further developed and promoted to give more visibility to the assessment and review of statistics in Scotland.' (Scottish Water)
- 'The proposals may go some way to improve public confidence in statistics, but this would need to be reviewed and further improvements identified on a regular basis.' (Clackmannanshire Council)
- 'We would welcome the involvement of the Parliament in the scrutiny of statistics and the ability of Parliamentary committees to question statisticians. We would suggest that this offer could be extended to those with statistical expertise in local authorities and other public bodies. We would welcome the opportunity to make our views known to the Parliament on relevant issues and to be questioned on what we do in relation to statistics.' (Falkirk Council)
- 'There is often an assumption that the audience for government statistics is statistically literate. This is not always the case. It would be useful for some work to be done on how statistics are presented and what they can be used for.' (Scottish Museums Council)
- 'Specifically, we consider it important that statistical releases and ministerial statements are separately branded and clearly differentiated. Ideally they should emanate from different locations, perhaps with some small time delay after the publication of the release before ministerial statement.' (Statistics Commission)
- 'Perhaps other independent experts (such as senior academic statisticians) could be employed to peer review statistics publications.' (West Dunbartonshire Council)
Finance Committee Scrutiny
74. The Finance Committee made the following relevant recommendations in its final report:
- Its successor committee takes evidence from the relevant Minister and the Scottish member of the proposed Statistics Board on the first annual report laid before Parliament and the proposed work programme of the Board. The Committee also suggests some form of ongoing scrutiny of the Board, certainly within the formative years following its creation.
- Its successor committee should be consulted on the draft Code of Practice. The Committee would therefore ask the Scottish Government, in their ongoing discussions with the UK Government on the implementation of the Bill, to ensure that the consultation on the draft code, and the draft code itself, be issued to its successor committee.
- The Scottish Government should give further consideration to the independence of the Chief Statistician and inform its successor Committee of the outcome of these deliberations.
The Annual Statistics Stakeholder Conference 2007
75. One of the workshops at the conference focussed on trust in Official Statistics. Key findings included:
- The vast majority of stakeholders trust Scottish official statistics in terms of the actual statistics in and off themselves as numbers. However, there is mistrust around the use of statistics by non-statisticians, particularly politicians and the media, where there is perception that they are misused and misinterpreted for non statistical purposes.
- There was a general sense that Scottish Official Statistics are accurate and robust in the main but that they are not entirely free from political interference.
- The important role of the National Statistics Code of Practice was cited.
- Some stakeholders felt that statistics are not free of politics because they are driven by the policy agenda which is subject to politics, performance indicators are political and can have a perverse effect on statistics, politicians can decided which statistics should be collected or not and politicians can 'cherry-pick' statistics to tell a biased story.
Scottish Government Response on Public Trust and Confidence:
76. Given the support from respondents we have developed and continue to develop the measures outlined at paragraph 72.
77. We have set up an internal working group of statistical staff to consider the branding and identity of Scottish Official Statistics. The focus of the early work of the group has been to make it clear that Scottish Official Statistics are produced by professionally independent statistical staff. The improvements that have been made as a result of this work were recently endorsed by the Statistics Commission in its final research report "Releasing Official Statistics - A Review of Statistical First Releases": 16
"We recommend that the approach taken in Scotland in Statistics News Releases which begin with the statement that "Scotland's Chief Statistician today published…. " should be adopted more generally to emphasise that official statistics are released by professional statisticians"
"Scottish Government practice is to issue a statistical news release, with a summary of main points, for all statistics. Political comment or ministerial statement, if any, then appears in a separate news release. This procedure seems to achieve a clear separation between the statistics and the political message, which some other departments do not manage."
78. We want to establish the confidence that our key stakeholders and commentators have in our statistics. The findings from the stakeholder event indicate that our users have higher levels of trust in official statistics than that of the general public. The new feedback form on the re-launched ScotStat website allows users to tell us their thoughts on any element of statistics. The ScotStat website and network will be furthered developed, with appropriate links to the UK Statistics Authority's assessment function.
79. Statisticians control the format, content, methodology and timing of statistical releases - this fact has to be better advertised to key commentators and the public. But equally we want to make it clear that Ministers decide on the scope or range of statistics produced, and that this is perfectly legitimate since they are democratically elected to pursue a policy agenda and need statistics to help them monitor progress. To suggest that this constitutes 'political interference' is misguided. The forthcoming Framework for Scottish Official Statistics will explain further the role of statisticians, Ministers and others in the management of government statistics.
80. As respondents to the consultation paper suggested, we must regularly monitor the impact that these changes are making - we are currently reviewing whether the change to our news releases has been relayed by the media to the public.
81. As well as reviewing the impact of such specific changes, we need to be able to measure and monitor public trust in statistics. Questions on public trust in official statistics were added to the Scottish Social Attitudes Survey for the first time in 2007. This research found that the most common reasons for expressing low levels of trust in the accuracy of government statistics were personal experience, the belief that statistics are misrepresented or 'spun' by politicians and the belief that statistics do not tell the whole story.
82. To have an impact on how the statistics relate to personal experience and to help to tell a more fuller story, we need to make it clearer how statistics are used in our publications, underline that it is the average position and where possible provide more analysis on different circumstances faced by the people of Scotland.
83. To change perceptions of misrepresentation or distortion of statistics we need help from Parliament, the media and the UK Statistics Authority - it is in everyone's interest that the statistical services provided by the government are trusted.
84. The UK Statistics Authority will report regularly to the Scottish Parliament, which should help raise the debate on statistical issues, and help underline that professionally independent statistical staff produce Official Statistics.